Physical activity slows femoral bone loss but promotes wrist fractures in postmenopausal women: a 15-year follow-up of the OSTPRE study.

Results on fracture risk among physically active persons are contradictory. The aim of this study was to investigate the long-term association between the self-reported physical activity (PA), the risk of fractures, and bone loss among peri- and postmenopausal women. The association between PA and fracture risk was examined during 15 years of follow-up in the population-based Osteoporosis Risk Factor and Prevention (OSTPRE) Study among 8560 women with a mean age of 52.2 years (range 47 to 56 years) at baseline. The amount and type of PA, as well as the types and mechanisms of fractures, were registered with self-administered questionnaires at 5-year intervals (ie, 1989, 1994, 1999, and 2004). A total of 2641 follow-up fractures were verified in 2073 women (24.2%). The study cohort was divided into quartiles by average hours of reported PA during the whole follow-up. Areal bone mineral density (aBMD) at the proximal femur (n = 2050) and lumbar spine (L(2)-L(4); n = 1417) was followed at 5-year intervals from a random stratified subsample with dual X-ray absorptiometry (DXA). Risk of fracture was estimated by using the Cox proportional hazards model with a mean follow-up time of 15.2 years. Weekly average time spent on leisure-time PA was 0.4, 1.7, 3.3, and 7.0 hours from the least to the most active quartiles, respectively. The risk of wrist fracture was higher in the active quartiles (II to IV) than in the most inactive quartile (I), with hazard ratios (HRs) of 1.3 [95% confidence interval (CI) 1.05-1.57, p = .014] for the second (II), 1.2 (95% CI 1.01-1.51, p = .045) for the third (III), and 1.4 (95% CI 1.14-1.69, p = .001) for the fourth (IV) quartile, respectively. Overall, most of the fractures were reported as a result of a fall (69.0%), with a 2.1 times higher rate of wrist fractures during the winter (November to April) than during summer season. There were no significant associations of PA with any other fracture types. Bone loss at the femoral neck, trochanter, and Ward's triangle was significantly associated with long-term PA (ANCOVA p < .05), whereas no associations of bone loss and PA in lumbar spine were seen. PA is associated with a moderate rise in wrist fracture risk, which might be explained in part by a higher number of outdoor activities. Regular PA of at least 1½ hours per week does not seem to increase the risk of other fractures and might significantly decrease proximal femur bone loss among peri- and postmenopausal women.